There was a front page link to the following article:
The study's conclusion was that drinking less on average led to faster marathon times.
Really? Examine how the study was set up: all study participants were to drink 8.5 oz. of water every 20 minutes.
The energy requirements of a marathon are mass x distance, which -- excluding variations in body weight of the runners -- are basically the same for faster and slower runners. So weight loss would nominally be the same. Weight gain will be dependent on how long they ran - 8.5 oz. more for every 20 minutes longer on the course. NET weight loss will thus be less for the slower runners.
I would not draw any interesting conclusions from this study.